open-consul/agent/consul/fsm/commands_oss.go

711 lines
26 KiB
Go
Raw Normal View History

package fsm
import (
"fmt"
"time"
2020-11-14 00:26:08 +00:00
"github.com/armon/go-metrics"
"github.com/armon/go-metrics/prometheus"
"github.com/hashicorp/consul/agent/consul/state"
"github.com/hashicorp/consul/agent/structs"
"github.com/hashicorp/consul/api"
)
2020-11-14 00:26:08 +00:00
var CommandsSummaries = []prometheus.SummaryDefinition{
{
Name: []string{"fsm", "register"},
2020-11-16 19:02:11 +00:00
Help: "Measures the time it takes to apply a catalog register operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "deregister"},
2020-11-16 19:02:11 +00:00
Help: "Measures the time it takes to apply a catalog deregister operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "kvs"},
2020-11-16 19:02:11 +00:00
Help: "Measures the time it takes to apply the given KV operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "session"},
2020-11-16 19:02:11 +00:00
Help: "Measures the time it takes to apply the given session operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "acl"},
2020-11-16 19:02:11 +00:00
Help: "Measures the time it takes to apply the given ACL operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "tombstone"},
2020-11-16 19:02:11 +00:00
Help: "Measures the time it takes to apply the given tombstone operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "coordinate", "batch-update"},
2020-11-16 19:02:11 +00:00
Help: "Measures the time it takes to apply the given batch coordinate update to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "prepared-query"},
2020-11-16 19:02:11 +00:00
Help: "Measures the time it takes to apply the given prepared query update operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "txn"},
2020-11-16 19:02:11 +00:00
Help: "Measures the time it takes to apply the given transaction update to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "autopilot"},
2020-11-16 19:02:11 +00:00
Help: "Measures the time it takes to apply the given autopilot update to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"consul", "fsm", "intention"},
Help: "Deprecated - use fsm_intention instead",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "intention"},
Help: "Measures the time it takes to apply an intention operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"consul", "fsm", "ca"},
Help: "Deprecated - use fsm_ca instead",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "ca"},
Help: "Measures the time it takes to apply CA configuration operations to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "ca", "leaf"},
Help: "Measures the time it takes to apply an operation while signing a leaf certificate.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "acl", "token"},
Help: "Measures the time it takes to apply an ACL token operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "acl", "policy"},
Help: "Measures the time it takes to apply an ACL policy operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "acl", "bindingrule"},
Help: "Measures the time it takes to apply an ACL binding rule operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "acl", "authmethod"},
Help: "Measures the time it takes to apply an ACL authmethod operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
{
Name: []string{"fsm", "system_metadata"},
Help: "Measures the time it takes to apply a system metadata operation to the FSM.",
2020-11-14 00:26:08 +00:00
},
// TODO(kit): We generate the config-entry fsm summaries by reading off of the request. It is
// possible to statically declare these when we know all of the names, but I didn't get to it
// in this patch. Config-entries are known though and we should add these in the future.
// {
// Name: []string{"fsm", "config_entry", req.Entry.GetKind()},
// Help: "",
// },
}
func init() {
registerCommand(structs.RegisterRequestType, (*FSM).applyRegister)
registerCommand(structs.DeregisterRequestType, (*FSM).applyDeregister)
registerCommand(structs.KVSRequestType, (*FSM).applyKVSOperation)
registerCommand(structs.SessionRequestType, (*FSM).applySessionOperation)
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// DEPRECATED (ACL-Legacy-Compat) - Only needed for v1 ACL compat
registerCommand(structs.ACLRequestType, (*FSM).applyACLOperation)
registerCommand(structs.TombstoneRequestType, (*FSM).applyTombstoneOperation)
registerCommand(structs.CoordinateBatchUpdateType, (*FSM).applyCoordinateBatchUpdate)
registerCommand(structs.PreparedQueryRequestType, (*FSM).applyPreparedQueryOperation)
registerCommand(structs.TxnRequestType, (*FSM).applyTxn)
registerCommand(structs.AutopilotRequestType, (*FSM).applyAutopilotUpdate)
registerCommand(structs.IntentionRequestType, (*FSM).applyIntentionOperation)
registerCommand(structs.ConnectCARequestType, (*FSM).applyConnectCAOperation)
registerCommand(structs.ACLTokenSetRequestType, (*FSM).applyACLTokenSetOperation)
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
registerCommand(structs.ACLTokenDeleteRequestType, (*FSM).applyACLTokenDeleteOperation)
registerCommand(structs.ACLBootstrapRequestType, (*FSM).applyACLTokenBootstrap)
registerCommand(structs.ACLPolicySetRequestType, (*FSM).applyACLPolicySetOperation)
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
registerCommand(structs.ACLPolicyDeleteRequestType, (*FSM).applyACLPolicyDeleteOperation)
registerCommand(structs.ConnectCALeafRequestType, (*FSM).applyConnectCALeafOperation)
registerCommand(structs.ConfigEntryRequestType, (*FSM).applyConfigEntryOperation)
registerCommand(structs.ACLRoleSetRequestType, (*FSM).applyACLRoleSetOperation)
registerCommand(structs.ACLRoleDeleteRequestType, (*FSM).applyACLRoleDeleteOperation)
registerCommand(structs.ACLBindingRuleSetRequestType, (*FSM).applyACLBindingRuleSetOperation)
registerCommand(structs.ACLBindingRuleDeleteRequestType, (*FSM).applyACLBindingRuleDeleteOperation)
registerCommand(structs.ACLAuthMethodSetRequestType, (*FSM).applyACLAuthMethodSetOperation)
registerCommand(structs.ACLAuthMethodDeleteRequestType, (*FSM).applyACLAuthMethodDeleteOperation)
wan federation via mesh gateways (#6884) This is like a Möbius strip of code due to the fact that low-level components (serf/memberlist) are connected to high-level components (the catalog and mesh-gateways) in a twisty maze of references which make it hard to dive into. With that in mind here's a high level summary of what you'll find in the patch: There are several distinct chunks of code that are affected: * new flags and config options for the server * retry join WAN is slightly different * retry join code is shared to discover primary mesh gateways from secondary datacenters * because retry join logic runs in the *agent* and the results of that operation for primary mesh gateways are needed in the *server* there are some methods like `RefreshPrimaryGatewayFallbackAddresses` that must occur at multiple layers of abstraction just to pass the data down to the right layer. * new cache type `FederationStateListMeshGatewaysName` for use in `proxycfg/xds` layers * the function signature for RPC dialing picked up a new required field (the node name of the destination) * several new RPCs for manipulating a FederationState object: `FederationState:{Apply,Get,List,ListMeshGateways}` * 3 read-only internal APIs for debugging use to invoke those RPCs from curl * raft and fsm changes to persist these FederationStates * replication for FederationStates as they are canonically stored in the Primary and replicated to the Secondaries. * a special derivative of anti-entropy that runs in secondaries to snapshot their local mesh gateway `CheckServiceNodes` and sync them into their upstream FederationState in the primary (this works in conjunction with the replication to distribute addresses for all mesh gateways in all DCs to all other DCs) * a "gateway locator" convenience object to make use of this data to choose the addresses of gateways to use for any given RPC or gossip operation to a remote DC. This gets data from the "retry join" logic in the agent and also directly calls into the FSM. * RPC (`:8300`) on the server sniffs the first byte of a new connection to determine if it's actually doing native TLS. If so it checks the ALPN header for protocol determination (just like how the existing system uses the type-byte marker). * 2 new kinds of protocols are exclusively decoded via this native TLS mechanism: one for ferrying "packet" operations (udp-like) from the gossip layer and one for "stream" operations (tcp-like). The packet operations re-use sockets (using length-prefixing) to cut down on TLS re-negotiation overhead. * the server instances specially wrap the `memberlist.NetTransport` when running with gateway federation enabled (in a `wanfed.Transport`). The general gist is that if it tries to dial a node in the SAME datacenter (deduced by looking at the suffix of the node name) there is no change. If dialing a DIFFERENT datacenter it is wrapped up in a TLS+ALPN blob and sent through some mesh gateways to eventually end up in a server's :8300 port. * a new flag when launching a mesh gateway via `consul connect envoy` to indicate that the servers are to be exposed. This sets a special service meta when registering the gateway into the catalog. * `proxycfg/xds` notice this metadata blob to activate additional watches for the FederationState objects as well as the location of all of the consul servers in that datacenter. * `xds:` if the extra metadata is in place additional clusters are defined in a DC to bulk sink all traffic to another DC's gateways. For the current datacenter we listen on a wildcard name (`server.<dc>.consul`) that load balances all servers as well as one mini-cluster per node (`<node>.server.<dc>.consul`) * the `consul tls cert create` command got a new flag (`-node`) to help create an additional SAN in certs that can be used with this flavor of federation.
2020-03-09 20:59:02 +00:00
registerCommand(structs.FederationStateRequestType, (*FSM).applyFederationStateOperation)
registerCommand(structs.SystemMetadataRequestType, (*FSM).applySystemMetadataOperation)
}
func (c *FSM) applyRegister(buf []byte, index uint64) interface{} {
defer metrics.MeasureSince([]string{"fsm", "register"}, time.Now())
var req structs.RegisterRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
// Apply all updates in a single transaction
if err := c.state.EnsureRegistration(index, &req); err != nil {
c.logger.Warn("EnsureRegistration failed", "error", err)
return err
}
return nil
}
func (c *FSM) applyDeregister(buf []byte, index uint64) interface{} {
defer metrics.MeasureSince([]string{"fsm", "deregister"}, time.Now())
var req structs.DeregisterRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
// Either remove the service entry or the whole node. The precedence
// here is also baked into vetDeregisterWithACL() in acl.go, so if you
// make changes here, be sure to also adjust the code over there.
if req.ServiceID != "" {
if err := c.state.DeleteService(index, req.Node, req.ServiceID, &req.EnterpriseMeta); err != nil {
c.logger.Warn("DeleteNodeService failed", "error", err)
return err
}
} else if req.CheckID != "" {
if err := c.state.DeleteCheck(index, req.Node, req.CheckID, &req.EnterpriseMeta); err != nil {
c.logger.Warn("DeleteNodeCheck failed", "error", err)
return err
}
} else {
if err := c.state.DeleteNode(index, req.Node); err != nil {
c.logger.Warn("DeleteNode failed", "error", err)
return err
}
}
return nil
}
func (c *FSM) applyKVSOperation(buf []byte, index uint64) interface{} {
var req structs.KVSRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "kvs"}, time.Now(),
[]metrics.Label{{Name: "op", Value: string(req.Op)}})
switch req.Op {
case api.KVSet:
return c.state.KVSSet(index, &req.DirEnt)
case api.KVDelete:
return c.state.KVSDelete(index, req.DirEnt.Key, &req.DirEnt.EnterpriseMeta)
case api.KVDeleteCAS:
act, err := c.state.KVSDeleteCAS(index, req.DirEnt.ModifyIndex, req.DirEnt.Key, &req.DirEnt.EnterpriseMeta)
if err != nil {
return err
}
return act
case api.KVDeleteTree:
return c.state.KVSDeleteTree(index, req.DirEnt.Key, &req.DirEnt.EnterpriseMeta)
case api.KVCAS:
act, err := c.state.KVSSetCAS(index, &req.DirEnt)
if err != nil {
return err
}
return act
case api.KVLock:
act, err := c.state.KVSLock(index, &req.DirEnt)
if err != nil {
return err
}
return act
case api.KVUnlock:
act, err := c.state.KVSUnlock(index, &req.DirEnt)
if err != nil {
return err
}
return act
default:
err := fmt.Errorf("Invalid KVS operation '%s'", req.Op)
c.logger.Warn("Invalid KVS operation", "operation", req.Op)
return err
}
}
func (c *FSM) applySessionOperation(buf []byte, index uint64) interface{} {
var req structs.SessionRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "session"}, time.Now(),
[]metrics.Label{{Name: "op", Value: string(req.Op)}})
switch req.Op {
case structs.SessionCreate:
if err := c.state.SessionCreate(index, &req.Session); err != nil {
return err
}
return req.Session.ID
case structs.SessionDestroy:
return c.state.SessionDestroy(index, req.Session.ID, &req.Session.EnterpriseMeta)
default:
c.logger.Warn("Invalid Session operation", "operation", req.Op)
return fmt.Errorf("Invalid Session operation '%s'", req.Op)
}
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// DEPRECATED (ACL-Legacy-Compat) - Only needed for legacy compat
func (c *FSM) applyACLOperation(buf []byte, index uint64) interface{} {
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// TODO (ACL-Legacy-Compat) - Should we warn here somehow about using deprecated features
// maybe emit a second metric?
var req structs.ACLRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "acl"}, time.Now(),
[]metrics.Label{{Name: "op", Value: string(req.Op)}})
switch req.Op {
case structs.ACLBootstrapInit:
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
enabled, _, err := c.state.CanBootstrapACLToken()
if err != nil {
return err
}
return enabled
case structs.ACLBootstrapNow:
2018-11-02 17:00:39 +00:00
// This is a bootstrap request from a non-upgraded node
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
if err := c.state.ACLBootstrap(index, 0, req.ACL.Convert(), true); err != nil {
return err
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// No need to check expiration times as those did not exist in legacy tokens.
if _, token, err := c.state.ACLTokenGetBySecret(nil, req.ACL.ID, nil); err != nil {
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
return err
} else {
acl, err := token.Convert()
if err != nil {
return err
}
return acl
}
case structs.ACLForceSet, structs.ACLSet:
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
if err := c.state.ACLTokenSet(index, req.ACL.Convert(), true); err != nil {
return err
}
return req.ACL.ID
case structs.ACLDelete:
return c.state.ACLTokenDeleteBySecret(index, req.ACL.ID, nil)
default:
c.logger.Warn("Invalid ACL operation", "operation", req.Op)
return fmt.Errorf("Invalid ACL operation '%s'", req.Op)
}
}
func (c *FSM) applyTombstoneOperation(buf []byte, index uint64) interface{} {
var req structs.TombstoneRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "tombstone"}, time.Now(),
[]metrics.Label{{Name: "op", Value: string(req.Op)}})
switch req.Op {
case structs.TombstoneReap:
return c.state.ReapTombstones(index, req.ReapIndex)
default:
c.logger.Warn("Invalid Tombstone operation", "operation", req.Op)
return fmt.Errorf("Invalid Tombstone operation '%s'", req.Op)
}
}
// applyCoordinateBatchUpdate processes a batch of coordinate updates and applies
// them in a single underlying transaction. This interface isn't 1:1 with the outer
// update interface that the coordinate endpoint exposes, so we made it single
// purpose and avoided the opcode convention.
func (c *FSM) applyCoordinateBatchUpdate(buf []byte, index uint64) interface{} {
var updates structs.Coordinates
if err := structs.Decode(buf, &updates); err != nil {
panic(fmt.Errorf("failed to decode batch updates: %v", err))
}
defer metrics.MeasureSince([]string{"fsm", "coordinate", "batch-update"}, time.Now())
if err := c.state.CoordinateBatchUpdate(index, updates); err != nil {
return err
}
return nil
}
// applyPreparedQueryOperation applies the given prepared query operation to the
// state store.
func (c *FSM) applyPreparedQueryOperation(buf []byte, index uint64) interface{} {
var req structs.PreparedQueryRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "prepared-query"}, time.Now(),
[]metrics.Label{{Name: "op", Value: string(req.Op)}})
switch req.Op {
case structs.PreparedQueryCreate, structs.PreparedQueryUpdate:
return c.state.PreparedQuerySet(index, req.Query)
case structs.PreparedQueryDelete:
return c.state.PreparedQueryDelete(index, req.Query.ID)
default:
c.logger.Warn("Invalid PreparedQuery operation", "operation", req.Op)
return fmt.Errorf("Invalid PreparedQuery operation '%s'", req.Op)
}
}
func (c *FSM) applyTxn(buf []byte, index uint64) interface{} {
var req structs.TxnRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSince([]string{"fsm", "txn"}, time.Now())
results, errors := c.state.TxnRW(index, req.Ops)
return structs.TxnResponse{
Results: results,
Errors: errors,
}
}
func (c *FSM) applyAutopilotUpdate(buf []byte, index uint64) interface{} {
var req structs.AutopilotSetConfigRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSince([]string{"fsm", "autopilot"}, time.Now())
if req.CAS {
act, err := c.state.AutopilotCASConfig(index, req.Config.ModifyIndex, &req.Config)
if err != nil {
return err
}
return act
}
return c.state.AutopilotSetConfig(index, &req.Config)
}
// applyIntentionOperation applies the given intention operation to the state store.
func (c *FSM) applyIntentionOperation(buf []byte, index uint64) interface{} {
var req structs.IntentionRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
// TODO(kit): We should deprecate this first metric that writes the metrics_prefix itself,
// the config we use to flag this out, telemetry.disable_compat_1.9 is on the agent - how do
// we access it here?
defer metrics.MeasureSinceWithLabels([]string{"consul", "fsm", "intention"}, time.Now(),
[]metrics.Label{{Name: "op", Value: string(req.Op)}})
defer metrics.MeasureSinceWithLabels([]string{"fsm", "intention"}, time.Now(),
[]metrics.Label{{Name: "op", Value: string(req.Op)}})
if req.Mutation != nil {
return c.state.IntentionMutation(index, req.Op, req.Mutation)
}
switch req.Op {
case structs.IntentionOpCreate, structs.IntentionOpUpdate:
connect: intentions are now managed as a new config entry kind "service-intentions" (#8834) - Upgrade the ConfigEntry.ListAll RPC to be kind-aware so that older copies of consul will not see new config entries it doesn't understand replicate down. - Add shim conversion code so that the old API/CLI method of interacting with intentions will continue to work so long as none of these are edited via config entry endpoints. Almost all of the read-only APIs will continue to function indefinitely. - Add new APIs that operate on individual intentions without IDs so that the UI doesn't need to implement CAS operations. - Add a new serf feature flag indicating support for intentions-as-config-entries. - The old line-item intentions way of interacting with the state store will transparently flip between the legacy memdb table and the config entry representations so that readers will never see a hiccup during migration where the results are incomplete. It uses a piece of system metadata to control the flip. - The primary datacenter will begin migrating intentions into config entries on startup once all servers in the datacenter are on a version of Consul with the intentions-as-config-entries feature flag. When it is complete the old state store representations will be cleared. We also record a piece of system metadata indicating this has occurred. We use this metadata to skip ALL of this code the next time the leader starts up. - The secondary datacenters continue to run the old intentions replicator until all servers in the secondary DC and primary DC support intentions-as-config-entries (via serf flag). Once this condition it met the old intentions replicator ceases. - The secondary datacenters replicate the new config entries as they are migrated in the primary. When they detect that the primary has zeroed it's old state store table it waits until all config entries up to that point are replicated and then zeroes its own copy of the old state store table. We also record a piece of system metadata indicating this has occurred. We use this metadata to skip ALL of this code the next time the leader starts up.
2020-10-06 18:24:05 +00:00
//nolint:staticcheck
return c.state.LegacyIntentionSet(index, req.Intention)
case structs.IntentionOpDelete:
connect: intentions are now managed as a new config entry kind "service-intentions" (#8834) - Upgrade the ConfigEntry.ListAll RPC to be kind-aware so that older copies of consul will not see new config entries it doesn't understand replicate down. - Add shim conversion code so that the old API/CLI method of interacting with intentions will continue to work so long as none of these are edited via config entry endpoints. Almost all of the read-only APIs will continue to function indefinitely. - Add new APIs that operate on individual intentions without IDs so that the UI doesn't need to implement CAS operations. - Add a new serf feature flag indicating support for intentions-as-config-entries. - The old line-item intentions way of interacting with the state store will transparently flip between the legacy memdb table and the config entry representations so that readers will never see a hiccup during migration where the results are incomplete. It uses a piece of system metadata to control the flip. - The primary datacenter will begin migrating intentions into config entries on startup once all servers in the datacenter are on a version of Consul with the intentions-as-config-entries feature flag. When it is complete the old state store representations will be cleared. We also record a piece of system metadata indicating this has occurred. We use this metadata to skip ALL of this code the next time the leader starts up. - The secondary datacenters continue to run the old intentions replicator until all servers in the secondary DC and primary DC support intentions-as-config-entries (via serf flag). Once this condition it met the old intentions replicator ceases. - The secondary datacenters replicate the new config entries as they are migrated in the primary. When they detect that the primary has zeroed it's old state store table it waits until all config entries up to that point are replicated and then zeroes its own copy of the old state store table. We also record a piece of system metadata indicating this has occurred. We use this metadata to skip ALL of this code the next time the leader starts up.
2020-10-06 18:24:05 +00:00
//nolint:staticcheck
return c.state.LegacyIntentionDelete(index, req.Intention.ID)
case structs.IntentionOpDeleteAll:
return c.state.LegacyIntentionDeleteAll(index)
case structs.IntentionOpUpsert:
fallthrough // unsupported
default:
c.logger.Warn("Invalid Intention operation", "operation", req.Op)
return fmt.Errorf("Invalid Intention operation '%s'", req.Op)
}
}
// applyConnectCAOperation applies the given CA operation to the state store.
func (c *FSM) applyConnectCAOperation(buf []byte, index uint64) interface{} {
var req structs.CARequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"consul", "fsm", "ca"}, time.Now(),
[]metrics.Label{{Name: "op", Value: string(req.Op)}})
defer metrics.MeasureSinceWithLabels([]string{"fsm", "ca"}, time.Now(),
[]metrics.Label{{Name: "op", Value: string(req.Op)}})
switch req.Op {
2018-04-07 00:58:45 +00:00
case structs.CAOpSetConfig:
if req.Config.ModifyIndex != 0 {
act, err := c.state.CACheckAndSetConfig(index, req.Config.ModifyIndex, req.Config)
if err != nil {
return err
}
return act
}
return c.state.CASetConfig(index, req.Config)
case structs.CAOpSetRoots:
act, err := c.state.CARootSetCAS(index, req.Index, req.Roots)
if err != nil {
return err
}
return act
case structs.CAOpSetProviderState:
act, err := c.state.CASetProviderState(index, req.ProviderState)
if err != nil {
return err
}
return act
case structs.CAOpDeleteProviderState:
if err := c.state.CADeleteProviderState(index, req.ProviderState.ID); err != nil {
return err
}
return true
case structs.CAOpSetRootsAndConfig:
act, err := c.state.CARootSetCAS(index, req.Index, req.Roots)
if err != nil {
return err
}
if !act {
return act
}
act, err = c.state.CACheckAndSetConfig(index+1, req.Config.ModifyIndex, req.Config)
if err != nil {
return err
}
return act
case structs.CAOpIncrementProviderSerialNumber:
sn, err := c.state.CAIncrementProviderSerialNumber(index)
if err != nil {
return err
}
return sn
default:
c.logger.Warn("Invalid CA operation", "operation", req.Op)
return fmt.Errorf("Invalid CA operation '%s'", req.Op)
}
}
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
// applyConnectCALeafOperation applies an operation while signing a leaf certificate.
func (c *FSM) applyConnectCALeafOperation(buf []byte, index uint64) interface{} {
var req structs.CALeafRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "ca", "leaf"}, time.Now(),
[]metrics.Label{{Name: "op", Value: string(req.Op)}})
switch req.Op {
case structs.CALeafOpIncrementIndex:
// Use current index as the new value as well as the value to write at.
// TODO(banks) do we even use this op any more?
if err := c.state.CALeafSetIndex(index, index); err != nil {
return err
}
return index
default:
c.logger.Warn("Invalid CA Leaf operation", "operation", req.Op)
return fmt.Errorf("Invalid CA operation '%s'", req.Op)
}
}
func (c *FSM) applyACLTokenSetOperation(buf []byte, index uint64) interface{} {
var req structs.ACLTokenBatchSetRequest
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "acl", "token"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "upsert"}})
opts := state.ACLTokenSetOptions{
CAS: req.CAS,
AllowMissingPolicyAndRoleIDs: req.AllowMissingLinks,
ProhibitUnprivileged: req.ProhibitUnprivileged,
Legacy: false,
FromReplication: req.FromReplication,
}
return c.state.ACLTokenBatchSet(index, req.Tokens, opts)
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
}
func (c *FSM) applyACLTokenDeleteOperation(buf []byte, index uint64) interface{} {
var req structs.ACLTokenBatchDeleteRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "acl", "token"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "delete"}})
return c.state.ACLTokenBatchDelete(index, req.TokenIDs)
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
}
func (c *FSM) applyACLTokenBootstrap(buf []byte, index uint64) interface{} {
var req structs.ACLTokenBootstrapRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "acl", "token"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "bootstrap"}})
return c.state.ACLBootstrap(index, req.ResetIndex, &req.Token, false)
}
func (c *FSM) applyACLPolicySetOperation(buf []byte, index uint64) interface{} {
var req structs.ACLPolicyBatchSetRequest
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "acl", "policy"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "upsert"}})
return c.state.ACLPolicyBatchSet(index, req.Policies)
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
}
func (c *FSM) applyACLPolicyDeleteOperation(buf []byte, index uint64) interface{} {
var req structs.ACLPolicyBatchDeleteRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "acl", "policy"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "delete"}})
return c.state.ACLPolicyBatchDelete(index, req.PolicyIDs)
New ACLs (#4791) This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week. Description At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers. On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though. Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though. All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management. Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are: A server running the new system must still support other clients using the legacy system. A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system. The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode. So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
2018-10-19 16:04:07 +00:00
}
func (c *FSM) applyConfigEntryOperation(buf []byte, index uint64) interface{} {
req := structs.ConfigEntryRequest{
Entry: &structs.ProxyConfigEntry{},
}
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
switch req.Op {
case structs.ConfigEntryUpsertCAS:
defer metrics.MeasureSinceWithLabels([]string{"fsm", "config_entry", req.Entry.GetKind()}, time.Now(),
[]metrics.Label{{Name: "op", Value: "upsert"}})
updated, err := c.state.EnsureConfigEntryCAS(index, req.Entry.GetRaftIndex().ModifyIndex, req.Entry)
if err != nil {
return err
}
return updated
case structs.ConfigEntryUpsert:
defer metrics.MeasureSinceWithLabels([]string{"fsm", "config_entry", req.Entry.GetKind()}, time.Now(),
[]metrics.Label{{Name: "op", Value: "upsert"}})
if err := c.state.EnsureConfigEntry(index, req.Entry); err != nil {
Centralized Config CLI (#5731) * Add HTTP endpoints for config entry management * Finish implementing decoding in the HTTP Config entry apply endpoint * Add CAS operation to the config entry apply endpoint Also use this for the bootstrapping and move the config entry decoding function into the structs package. * First pass at the API client for the config entries * Fixup some of the ConfigEntry APIs Return a singular response object instead of a list for the ConfigEntry.Get RPC. This gets plumbed through the HTTP API as well. Dont return QueryMeta in the JSON response for the config entry listing HTTP API. Instead just return a list of config entries. * Minor API client fixes * Attempt at some ConfigEntry api client tests These don’t currently work due to weak typing in JSON * Get some of the api client tests passing * Implement reflectwalk magic to correct JSON encoding a ProxyConfigEntry Also added a test for the HTTP endpoint that exposes the problem. However, since the test doesn’t actually do the JSON encode/decode its still failing. * Move MapWalk magic into a binary marshaller instead of JSON. * Add a MapWalk test * Get rid of unused func * Get rid of unused imports * Fixup some tests now that the decoding from msgpack coerces things into json compat types * Stub out most of the central config cli Fully implement the config read command. * Basic config delete command implementation * Implement config write command * Implement config list subcommand Not entirely sure about the output here. Its basically the read output indented with a line specifying the kind/name of each type which is also duplicated in the indented output. * Update command usage * Update some help usage formatting * Add the connect enable helper cli command * Update list command output * Rename the config entry API client methods. * Use renamed apis * Implement config write tests Stub the others with the noTabs tests. * Change list output format Now just simply output 1 line per named config * Add config read tests * Add invalid args write test. * Add config delete tests * Add config list tests * Add connect enable tests * Update some CLI commands to use CAS ops This also modifies the HTTP API for a write op to return a boolean indicating whether the value was written or not. * Fix up the HTTP API CAS tests as I realized they weren’t testing what they should. * Update config entry rpc tests to properly test CAS * Fix up a few more tests * Fix some tests that using ConfigEntries.Apply * Update config_write_test.go * Get rid of unused import
2019-04-30 23:27:16 +00:00
return err
}
return true
case structs.ConfigEntryDelete:
defer metrics.MeasureSinceWithLabels([]string{"fsm", "config_entry", req.Entry.GetKind()}, time.Now(),
[]metrics.Label{{Name: "op", Value: "delete"}})
return c.state.DeleteConfigEntry(index, req.Entry.GetKind(), req.Entry.GetName(), req.Entry.GetEnterpriseMeta())
default:
return fmt.Errorf("invalid config entry operation type: %v", req.Op)
}
}
func (c *FSM) applyACLRoleSetOperation(buf []byte, index uint64) interface{} {
var req structs.ACLRoleBatchSetRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "acl", "role"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "upsert"}})
return c.state.ACLRoleBatchSet(index, req.Roles, req.AllowMissingLinks)
}
func (c *FSM) applyACLRoleDeleteOperation(buf []byte, index uint64) interface{} {
var req structs.ACLRoleBatchDeleteRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "acl", "role"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "delete"}})
return c.state.ACLRoleBatchDelete(index, req.RoleIDs)
}
func (c *FSM) applyACLBindingRuleSetOperation(buf []byte, index uint64) interface{} {
var req structs.ACLBindingRuleBatchSetRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "acl", "bindingrule"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "upsert"}})
return c.state.ACLBindingRuleBatchSet(index, req.BindingRules)
}
func (c *FSM) applyACLBindingRuleDeleteOperation(buf []byte, index uint64) interface{} {
var req structs.ACLBindingRuleBatchDeleteRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "acl", "bindingrule"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "delete"}})
return c.state.ACLBindingRuleBatchDelete(index, req.BindingRuleIDs)
}
func (c *FSM) applyACLAuthMethodSetOperation(buf []byte, index uint64) interface{} {
var req structs.ACLAuthMethodBatchSetRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "acl", "authmethod"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "upsert"}})
return c.state.ACLAuthMethodBatchSet(index, req.AuthMethods)
}
func (c *FSM) applyACLAuthMethodDeleteOperation(buf []byte, index uint64) interface{} {
var req structs.ACLAuthMethodBatchDeleteRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
defer metrics.MeasureSinceWithLabels([]string{"fsm", "acl", "authmethod"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "delete"}})
return c.state.ACLAuthMethodBatchDelete(index, req.AuthMethodNames, &req.EnterpriseMeta)
}
wan federation via mesh gateways (#6884) This is like a Möbius strip of code due to the fact that low-level components (serf/memberlist) are connected to high-level components (the catalog and mesh-gateways) in a twisty maze of references which make it hard to dive into. With that in mind here's a high level summary of what you'll find in the patch: There are several distinct chunks of code that are affected: * new flags and config options for the server * retry join WAN is slightly different * retry join code is shared to discover primary mesh gateways from secondary datacenters * because retry join logic runs in the *agent* and the results of that operation for primary mesh gateways are needed in the *server* there are some methods like `RefreshPrimaryGatewayFallbackAddresses` that must occur at multiple layers of abstraction just to pass the data down to the right layer. * new cache type `FederationStateListMeshGatewaysName` for use in `proxycfg/xds` layers * the function signature for RPC dialing picked up a new required field (the node name of the destination) * several new RPCs for manipulating a FederationState object: `FederationState:{Apply,Get,List,ListMeshGateways}` * 3 read-only internal APIs for debugging use to invoke those RPCs from curl * raft and fsm changes to persist these FederationStates * replication for FederationStates as they are canonically stored in the Primary and replicated to the Secondaries. * a special derivative of anti-entropy that runs in secondaries to snapshot their local mesh gateway `CheckServiceNodes` and sync them into their upstream FederationState in the primary (this works in conjunction with the replication to distribute addresses for all mesh gateways in all DCs to all other DCs) * a "gateway locator" convenience object to make use of this data to choose the addresses of gateways to use for any given RPC or gossip operation to a remote DC. This gets data from the "retry join" logic in the agent and also directly calls into the FSM. * RPC (`:8300`) on the server sniffs the first byte of a new connection to determine if it's actually doing native TLS. If so it checks the ALPN header for protocol determination (just like how the existing system uses the type-byte marker). * 2 new kinds of protocols are exclusively decoded via this native TLS mechanism: one for ferrying "packet" operations (udp-like) from the gossip layer and one for "stream" operations (tcp-like). The packet operations re-use sockets (using length-prefixing) to cut down on TLS re-negotiation overhead. * the server instances specially wrap the `memberlist.NetTransport` when running with gateway federation enabled (in a `wanfed.Transport`). The general gist is that if it tries to dial a node in the SAME datacenter (deduced by looking at the suffix of the node name) there is no change. If dialing a DIFFERENT datacenter it is wrapped up in a TLS+ALPN blob and sent through some mesh gateways to eventually end up in a server's :8300 port. * a new flag when launching a mesh gateway via `consul connect envoy` to indicate that the servers are to be exposed. This sets a special service meta when registering the gateway into the catalog. * `proxycfg/xds` notice this metadata blob to activate additional watches for the FederationState objects as well as the location of all of the consul servers in that datacenter. * `xds:` if the extra metadata is in place additional clusters are defined in a DC to bulk sink all traffic to another DC's gateways. For the current datacenter we listen on a wildcard name (`server.<dc>.consul`) that load balances all servers as well as one mini-cluster per node (`<node>.server.<dc>.consul`) * the `consul tls cert create` command got a new flag (`-node`) to help create an additional SAN in certs that can be used with this flavor of federation.
2020-03-09 20:59:02 +00:00
func (c *FSM) applyFederationStateOperation(buf []byte, index uint64) interface{} {
var req structs.FederationStateRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
switch req.Op {
case structs.FederationStateUpsert:
defer metrics.MeasureSinceWithLabels([]string{"fsm", "federation_state", req.State.Datacenter}, time.Now(),
[]metrics.Label{{Name: "op", Value: "upsert"}})
if err := c.state.FederationStateSet(index, req.State); err != nil {
return err
}
return true
case structs.FederationStateDelete:
defer metrics.MeasureSinceWithLabels([]string{"fsm", "federation_state", req.State.Datacenter}, time.Now(),
[]metrics.Label{{Name: "op", Value: "delete"}})
return c.state.FederationStateDelete(index, req.State.Datacenter)
default:
return fmt.Errorf("invalid federation state operation type: %v", req.Op)
}
}
func (c *FSM) applySystemMetadataOperation(buf []byte, index uint64) interface{} {
var req structs.SystemMetadataRequest
if err := structs.Decode(buf, &req); err != nil {
panic(fmt.Errorf("failed to decode request: %v", err))
}
switch req.Op {
case structs.SystemMetadataUpsert:
defer metrics.MeasureSinceWithLabels([]string{"fsm", "system_metadata"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "upsert"}})
if err := c.state.SystemMetadataSet(index, req.Entry); err != nil {
return err
}
return true
case structs.SystemMetadataDelete:
defer metrics.MeasureSinceWithLabels([]string{"fsm", "system_metadata"}, time.Now(),
[]metrics.Label{{Name: "op", Value: "delete"}})
return c.state.SystemMetadataDelete(index, req.Entry)
default:
return fmt.Errorf("invalid system metadata operation type: %v", req.Op)
}
}